Seismic evidence for Mesozoic sedimentary troughs on the. Hebridean continental margin. lONES'S interpretation of seismic refrac- tion lines on the NW British ...
Nature Vol. 276
23 Nm'cmber 1978
matters arising Seismic evidence for Mesozoic sedimentary troughs on the Hebridean continental margin lONES'S interpretation of seismic refraction lines on the NW British continental shelf l provides welcome corroboration of the Ml ,zoic sedimentary basin between the FI"I.nan Isles and Lewis first postulated by Bullerwell (see refs 2-4), As we have already suggested that the Rockall Trough developed by seafloor spreading in the early Permian 5 - 7 , we wish we could also welcome lones' interpretation of the structure of the NW British margin, which su~gests that the development of the Trough, at least in part, goes back to Permo-Triassic times, However, we interpret the high velocity refractor (Vp = 4.4 ± 0.3 km S-I) as corresponding merely to the top of a layer of PalaeoceneEocene basalts, because (1) regional mapping and study of interval velocities using commercial 24-fold reflection lines and aeromagnetic coverage shows that basalts extend southwards from the Faeroes over almost the whole of the Rockall Trough north of Anton Dohrn seamount (as was suggested by Roberts 8 ) and also over the outer NW Scottish shelf from the Wyville-Thomson Ridge to about 58° N; (2) the high-frequency negative magnetic anomaly field characteristic of these basalts at shallow depth is clearly seen both on lones' magnetic profile l and on the IGS aeromagnetic map2; and (3) the refractor velocity is similar to that for basalis elsewhere on the NW European margin 9 - II , but probably too high for Permo-Triassic sediments; although the interval velocity for pure Permo--Triassic sandstones at -4 km depth in wells in the Viking Graben l2 and West Shetland Basin 13 can exceptionally attain values as high as 4.5 km S-I, the usual range of velocity is 3.6 - 4.1 km S-I for Permo-Triassic sands tones in these areas. The Permo-Triassic beneath the Little Minch has a Vp - 3,9 km S-I .. obtained by refraction shooting l4 . Above the basalts, the refractors with apparent Vp - 2.8-3.1 km S-I could be due to Quaternary glacial till; measurements on data from the West Shetland area show that glacial till up to 100 m thick can have interval velocities V;n,- 2.6-2,9 km S-I. However, the Tertiary sediments between the Quaternary and the basalts have a lower Vin " so the top of the basalts is somewhat shallower than that given by lones' profiles. The gravity suggests that there may be sediments beneath the basalts, but we 0028 ~836!18/0276-O420$01.00
have no way of estimating their age. Thus we conclude that lones' results have no relevance to the early evolution of the Rockall Trough, and, furthermore, his attempt to estimate the hydrocarbon prospects of this part of the margin is misleading; prospects here are poorer than elsewhere, because of the basalt cover. However, our reconstruction' suggests that in mid-lurassic times Orphan Knoll was situated at the southeast margin of the Rockall Trough. The neritic Bajocian sands found beneath the Knoll 15 may therefore occur elsewhere along the margins of the Trough, and may also overlie downfaulted basins of Carboniferous age on the outer continental margin, developed during an intracontinental rifting phase 6 before early Permian spreading. Therefore along the margins clear of Tertiary basalt cover, for example, the south-east Faeroe-Shetland Trough margin between 60 and 62° N, and both margins of the Rockall Trough south of about 58° N, the hydrocarbon prospects may be very gQod, D. K. SMYTHE N. KENOLTY Institute of Geological Sciences, Marine Geophysics Unit, MurchisonHouse, West Mains Road, Edinburgh, UK M. 1. RUSSELL Department of Applied Geology, University of Strathclyde, lames Weir Building, 75 Montrose Street, Glasgow, UK 1. Jon... E. J. W. Nature 272, 789-792 (1978). 2. Bullerwell, W. Aeromagnt'tic Map of Parr of G,eat Britain and Northern Ireland. Shut 12 (GeoL Surv. G.B .. 1968). 3. Eden. R. A .. Wriaht. J. E. & Bullerwell. W. Rep. Inst. geol. Sci. No. 70/14, 111-128 (1971). 4. Dunham. K. The Sub-Pleistocene Geology of the Brirish Islts and Adjacent Continental Shelf (Inst. geol. Sci., 1972). 5. Russell. M. J. in implicah·ons of Contint"nlai Drifl to th~ Earth Sciences (t'ds Tarling. D. H. & Runcorn. S. K.) 581-597 (Academic. London, 1973). 6. Ru"ell. M. J. Scott. 1. Gwl. 12,315-323 (1976). 7. Ruell. M. 1. & Smythe, D. K. in Pmology and Geachemisrry of Continental Rifts (eds Neumann. E. R. & Ramherg. I. B.) 173179 (Reidel. Dordrecht. 1978). 8. Roberts. D. G. Dap-Sea Res. 18,353 -360 (1971). 9. palmason. G. TFcwnophy.,ics. 2,475 -482 (1965). 10. Chalmers. J. A .. Dobin,on. A .. Mould, A & Smythe. D. K. Geophys.1. R. astr. Soc. 49, 288 (1977). 11. Talwani. M. & Eldholm. O. Bull. ieol Soc. Am. 83, 3575-3606 (1972). 12. Kent, P. E. J. geol. Soc. Lond. 131,435-468 (1975). 13. Cashion. W. W. in Offshore EUrope 75. Paper OE-75·216 (Spearhead. Kin&S,on-upon- Thames. 1975). 14. Smythe. D. K .• Sowerbutt •• W. T. C. Bacon. M. & McOuillin. R. Nature phys. Sci. 236, 87 .. 89 (1972). 15. Laughton. A. S. et al. in Init. Rep. DSDP Leg 12 33-159 (1972).
lONES REPLIES-The uncertainties involved in estimating stratigraphical ages from seismic velocities in the Hebridean region were emphasised in my paperl and
I am therefore grateful to Smythe et al. for their early comments. Although these are primarily based on their unpublished reflection profiles, velocity data and regional maps, some further pertinent remarks can be made. Smythe et al. accept my conclusion that Mesozoic sediments occur immediately west of Lewis, an area in which the age of the sediment cover was previollsly undefined 2. My inference that Mesozoic sediments also lie close to the sea floor north-west of the F1annan Ridge is based on the observation that the 2.8-3.1 km S-I values on lines lM-9 and lM-lOB are significantly higher than published velocities in Tertiary-Quaternary sequences l . Thus constrained, I was unable to interpret the deeper 4.0-4.4 km S-I refractor as the top of a pile of Tertiary basaltic lavas. Using 'measurements on data' from an unspecified location near the Shetlands, Smythe et al. attribute the 2.83.1 km S-I values to glacial till, which then allows them to suggest that the 4.04.4 km S-I layer is much younger thar I proposed. As they do not indicate the water depths in which the high Pleistoccne velocities were d.:termined, the method of measurement and also local velocity changes, it is not clear whether a direct comparison with shallow refractor velocities near the F1annan Isles is ju~tified. Are their 2.6-2,9 km s -I values sufficiently typical of the Pleistocene to permit an extrapolation over a large area? Their results may be of regional importance but as careful mapping by Eden 3 and others has revealed marked and often rapid lateral variations in Pleistocene deposits off northern Britain, Smythe et at. need to provide additional evidence to show that thick, high velocity boulder clay is likely to lie near the shelf break, some 300 km south-west of the Shetlands. They must also account in their interpretation for the presence of preQuaternary reflectors 4.s within the westwards continuation of their proposed Quaternary sequence .. In their discussion of the 4,0-4.4 km S-I refractor, four references are misleadingly quoted (their refs 10-13) to add weight to their contention that it represents Tertiary basaltic lavas and not pre-lurassic sediments. Seismic velocities for basalts are omitted in ref. 10. Talwani and Eldholm l1 ascribe velocities near 4.4 km S-1 on the Norwegian margin, not to basalts, but to pre-Cretaceous sediments. A limiting 4.5 km S-1 for Permo-Triassic sediments is not included in their refs 12, 13. My values of 4.0-4.4 km S-I seem to be well ©
Macmillan Journals Lld 1'178
Nature Vol. 276
23 November 1978
within the permissible range for PermoTrias as the refraction profiles of Browitt 6 show velocities of 4.6-4.8 km S-1 at a level in the West Shetland Basin where Cashion's sections 7 reveal thick sediments of this age. Thus, pre-Jurassic deposits cannot be dismissed at the depths I have indicated. They correctly note that magnetic anomalies occur over the 4.04.4 km S-1 layer but, rather than arising from Tertiary lavas, these can be interpreted as reflecting small basic intrusions within a thick Mesozoic section and magnetisation contrasts in the Precambrian basement. In denying that my results are relevant to the early evolution of the Rockall Trough, Smythe et al. have ignored my demonstration of major faulting and crustal subsidence. As they accept a Mesozoic age for the Flannan T, ough they must also accept the existence on the continental margin of NNE-trending faults which were active during Mesozoic time. As these run approximately parallel to the continental slope north of St Kilda and have increasingly larger displacements northwestwards\ it is difficult to resist the conclusion that the faulting is directly associated with the major phase of down warping in the Rockall Trough. My section showing thick pre-Jurassic sediments north-west of the fault-bounded Flannan Ridge is compatible with the Permian opening first suggested by Heirtzler and Hayes B but it does not establish that seafloor spreading took place at this time. Commenting on petroleum prospects off northern Britain, Smythe et al. offer a comparison with other parts of the Rockall Trough that hinges on unpublished data and a geometrical argument which is questionable because the extent of oceanic crust in the region is unknown. Several parts of the Rockall Trough clearly deserve detatled commercial expl"ration, although my appraisal of prospects would be less sweeping than that given by Smythe et al. Moreover, in view of the geophysical indications of thick sediments and major Mesozoic faulting west of Hebrides, it seems premature for them to take a firm, pessimistic stand on a large section of the Scottish margin, especially since drilling investigations are at such an early stage. E. J. W. JONES • Department of Geology, University College, Gower St, London WC1, UK I. Jones. E. J. W. Natu" 272, 789-792 (1978). 2. Dunham. K. The Sub-Pleistoane Geology of the Bn'tish Isles and the Adjacent Continental Shelf (lnst. aeol. Sci., 1972). 3. Eden. R. A .• Holmes. R. & Fannin. N. G. T. Rep. Iltst. etol. Sci. No. 77/15 (1978). 4. Jones. E. J. W .. Ewing. M .• Ewing. J. I. & Eitlreim, S. L. J. erophys. Rn 75,1655-1680 (1970). 5. Ruddiman, W: F. Geol. Soc. Am. Bull. 83, 2039-2062 (1972). 6. Browitt. C W. A. Nature 236, 161-163 (1972). 7. Cashion, W. W. in Offshore Europe 75, Paper OE· 75·216 (Spearhead. Kingston·upon· Thames. 1975). 8. Heirtzler. J. R. & Hayes, D. E. Sciena 157, 185-187 (1967).
Object reconstruction from intensity data GULL AND DANIELLI have recently proposed a new maximum entropy algorithm for the reconstruction of an object field from incomplete and noisy measurements of its Fourier transform, This new algorithm may encourage other workers to apply maximum entropy techniques to their own problems. For this reason we point out here that the application of maximum entropy techniques for object reconstruction in the absence of phase information could lead to wholly false conclusions due to ambiguities. We shall illustrate some of the problems using a specific example. Many such examples may easily be generated and the one chosen here is not unusual. Figure 1 shows four different, real and positive objects which all give rise to the scattered intensity shown in Fig. 1 e, Families of dissimilar objects whose Fourier transforms have the same modulus occur whenever the scattered intensity has non-zero minima 2 , Such minima indicate the presence of complex zeros in the scattered field. For each such zero there are two possible positions, which correspond to different objects whose Fourier transforms have identical moduli. The possible positions for these zeros are symmetrical with respect to the real axis
Fig. 1 All four different real positive objects shown in a, b, c and d produce the same far field intensity, shown in e. The four different objects have been generated by reflecting or 'flipping' the zeros of F(z) about the x-axis, as described in the text.
and thus one speaks of 'flipping' zeros about this axis in order to generate different objects, The example in Fig, 1 was produced by 'flipping' such zeros, Clearly the four objects thus created are very different, for example, in the number and position of their peaks, The maximum entropy algorithm should converge to one of these possible solutions but the only criterion it has as a basis for the selection of that solution is that of smoothness, as all the possibilities are real and positive. However, inspection of the four possible solutions shows that none may be disregarded as unphysicalthey are all 'smooth', It would only be possible to identify one of these possibilities as the solution if one has some extra information, such as approximate phase values which may provide a basis for such discrimination. On the basis of the intensity alone there can be no unique solution, unless all minima are zero. The selection of one solution using